P3‐111: Cognitive assessment: Discrimination of impairment and detection of decline in Alzheimer's disease and mild cognitive impairment

Abstract

Traditional outcomes used in clinical trials of Alzheimer's disease (AD) are thought to lack sensitivity to detect subtle changes in cognition. Tests of specific cognitive domains may be more sensitive, allowing for earlier detection of cognitive change in clinical trials. Cognitive tests targeting memory (episodic, working, semantic and spatial), processing speed, and executive function were administered to 30 healthy elderly (HE, mean 79 years), 17 Mild Cognitive Impairment (MCI, mean 82 years) and 12 AD (mean 83 years) at baseline, 6 months, and 1 year. The objective was to characterize the cross-sectional and longitudinal performance of the tests by assessing (1) their ability to discriminate across levels of impairment, and (2) the presence of ceiling or floor effects and (3) sensitivity to cognitive decline over time. Tests were correlated with the Everyday Memory Questionnaire (EMQ), a measure of functional memory loss. Significant differences in group means across 3 groups noted. Tests of episodic (CANTAB PAL), semantic (GNT), and spatial (route learning and recall) memory and executive function discriminated performance across HE, MCI, AD (p<0.05) better than episodic memory (CAMCOG learning), processing speed or working memory. Episodic and spatial memory scores were strongly correlated with EMQ, supporting construct validity of these measures (Pearson r ranged from 0.45–0.56). Episodic memory scores for MCI and AD trended downward at 6 months, but no significant changes were detected. Results suggest that some targeted tests of memory and executive function may be more responsive than global cognitive measures in dementia trials.